Abstract DBS therapy for Parkinson's disease is now the primary surgical approach for Parkinson's disease, recently FDA approved at 4 years after onset of disease. However, this therapy is still limited to treatment of a subset of motor symptoms (ie, tremor, rigidity, bradykinesia and dyskinesias) and requires considerable postoperative clinical adjustment to program and maintain function. A number of improvements to DBS for PD are being tested, including changes in patterns of stimulation and specific targets. However, a major new approach involves internal parameter adjustment using a surrogate physiological marker of clinical symptoms, useful for confirming initial electrode placement, programming, and also long-term optimization of parameters. Several research systems have been suggested and are in testing for development of closed loop systems, including systems based on recording beta-band oscillations. Closed loop control involving recording a surrogate signal relevant to PD could improve DBS therapy on several time scales, including short-term dynamics (ie, over minutes), initial programming (over weeks to months), and long-term, depending on the time course of response to STN DBS. In addition to spontaneous beta band recording we have also implemented direct evoked potential recording using the stimulating DBS electrode, requiring suppression of the DBS-induced artifact. These intraoperative DBS recordings during STN DBS implants have revealed a complex evoked potential likely reflecting GPe/GPi activation, which may provide an excellent surrogate marker. This complex evoked potential changes over a short-term time period as the treatment effect of STN DBS comes on, indicating that the evoked potential likely reflects DBS effects on a larger motor circuit as the circuit dynamically is altered to an improved state. We hypothesize that this surrogate marker (in addition to beta band oscillations) may provide a key feedback signal for scalar, graded (proportional) closed loop DBS control, highly relevant to DBS effects on PD circuitry. To test this hypothesis we will perform long-term recording of this signal from humans (in either STN or GPe/GPi) together with DBS stimulation (in STN and/or GPi), using a novel DBS recording/stimulation device (Medtronics RC+S). These clinical experiments will focus on a small, pilot clinical study (n = 6 patients) to implant bilateral STN + GPe/GPi DBS electrodes in Parkinson's patients eligible for DBS using conventional stereotactic localization, connecting to Medtronics RC+S IPGs. Patients will benefit from either ordinary STN or GPi DBS stimulation and then we will also test the possibility of synergism between the two electrodes for clinical efficacy. Additionally, we will analyze the motor efficacy of both an external (using recording and modifying the parameters manually) and internal (using an algorithm for providing parameters automatically) scalar, closed loop response to these recorded surrogate markers. We will take advantage of the graded nature of the spontaneous and evoked responses to construct a proportional control feedback system, and to specifically delineate the time constants of the closed loop system to be able to define optimally damped control of PD symptoms. These experiments will provide a number of novel outcomes, including a direct, within-person comparison of STN and GPe/GPi DBS efficacy, development of an optimal surrogate parameter for detecting DBS efficacy using spontaneous and evoked physiological responses in direct comparison to clinical symptoms, and defining an optimal, scalar feedback, proportional control system for treatment on various time scales.